The impact of technological related M&A’s on the innovative performance
of a firm: Evidence from High Tech and Low-Tech Industries
Master Thesis
August 5, 2017
Leidy Bedoya
Student Number: 11414413
MSc. Business Administration-Strategy Track
University of Amsterdam
Statement of Originality
This document is written by Leidy Bedoya who declares to take full responsibility for the contents of this document. I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it. The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.
Abstract
This paper takes on a resource based knowledge view to analyze the phenomenon of M&A and Innovation: firms undergo M&A’s as a way to acquire new resources and capabilities and enhance innovation. In this paper, I will investigate the differential effects of acquisitions in the Low-Tech and High-Tech industries on the innovative performance of a firm. In particular, this paper tries to verify four hypotheses. First, it investigates whether the number of tech acquisitions is positively associated with the total innovative performance of a firm. Secondly, this research examines whether the technological relatedness positively affects the post exploitative innovation performance of a firm and whether the technological relatedness is non-significantly or negatively associated with the post exploratory innovation performance. Finally, the paper identifies if the R&D intensity of the acquirer firm moderates the relationship between the number of technological acquisitions and the exploratory innovation performance of the firm.
Keywords: M&A; innovative performance; exploitation; exploration; High-Tech and Low-Tech; Technological Relatedness; R&D
Table of Contents
1. Introduction ... 5
2. Literature Review ... 8
3. Research Question and Hypotheses ... 12
3.2 Theoretical background and hypotheses ... 14
4. Data and Methods ... 18
4.1. Model ... 18
4.2 Data and Samples ... 19
4.3 Variables ... 21
4.3.1 Independent Variables ... 21
4.3.2 Dependent variables ... 24
4.3.3 Control and moderating variables ... 25
5. Results ... 26
5.2 Regression Results and Conclusions ... 30
6. Discussion ... 38
6.1 Limitations and Further Research ... 41
Reference List ... 42
List of Figures Conceptual models ..………..…….14
List of Tables 1 Sample Descriptive Statistics……….….…….28
2 Description of Size Quartiles………...………28
3 Correlation Matrices……….………...29
4 Regression Results Hypothesis 1……….32
5 Regression Results Hypothesis 2……….33
6 Regression Results Hypothesis 3……….36
1. Introduction
Firms perform Mergers and Acquisitions (M&A) for a multitude of reasons such as entering new markets, acquiring strategic resources, learning and strengthening their market positioning (Ahuja and Novelli, 2014). To better understand whether acquisitions have a positive or negative effect on organizations, authors have analyzed the relationship of mergers and acquisitions and diverse dimensions of performance. Some of these dimensions are profitability, employment, labor productivity, innovativeness and others. These studies have researched different aspects of mergers and acquisitions like the antecedents for acquisitions, the internal and external factors that moderate acquisition performance and other type of acquisition outcomes (Haleblian et al, 2009).
This thesis will focus specifically on the M&A-Innovation relationship from a Resource Based View (RBV) perspective. From the RBV perspective, acquisitions are seen as a way for the acquirer firm to acquire new resources and capabilities and enhance innovation (Ahuja and Novelli, 2014). This is due to the fact that by acquiring a firm, the acquirer firm gets access to a new range of technological resources, capabilities, complementary resources and commercialization competences. These new acquired resources not only make it possible for the acquirer to enhance the number of elements for recombination, but it also enables the acquiring firm to use its current resources in a different way which may lead to a higher innovation output (Kaul, 2012; Ahuja and Novelli, 2014).
There has been extensive research done on the M&A RBV logic aspect and scholars have focused on different angles of the relationship between M&A’s and Innovation. Some of the angles researched in previous literature on this topic are how the characteristics of the knowledge and/or technological resources of the acquirer/target firms increase the likelihood of a firm engaging in an M&A. For example, in their research Bena and Kai Li (2014) show
that firms with low patent portfolios and high research and development expenses tend to be the targets in an M&A transaction. Additionally, their study also shows that technological overlap between two firms has a positive effect on transaction incidence.
Other authors have focused on researching how different characteristics of acquirers/target firms can lead to an improvement in the innovation performance of a firm (e.g. Cloodt et al., 2006; Porrini, 2004; Karim and Mitchell, 2000). Previous research on this topic has shown that technological acquisitions in contrast to technological acquisitions have a negative or non-significant impact on the post M&A innovation output of the acquirer firm (Ahuja and Katila, 2001; Cloodt et al, 2001; Wagner, 2011).
Previous literature researching the relationship between technological or non-technological acquisitions and the post innovative output of the firm has mostly focused on firms involved in the high-tech industries. However, this relationship has remained unexplored for firms involved in the low-tech industries. The main reason for this has been that many of these studies have chosen patent counts to operationalize the innovation output of firms. Patents are overall a good measure for innovation, but there are some limitations on the use of patents as a measurement for the total innovation output of the firm. For instance, some innovations are not patentable or are maybe not patented. Previous literature and appropriability logic indicate that these problems can vary significantly across industries (Levin et al., 1987; Cohen and Levinthal 1989; Ahuja and Katila, 2001). So, in order to minimize the factors that affect patenting propensity, authors have chosen to limit their studies to only a few closely related sectors. (Ahuja and Katila, 2001; Cloodt et al, 2001; Wagner 2011).
High tech and low-tech industries have different characteristics. Firms operating in a high-tech industry tend to use acquisitions to diversify their technology, to gain quick access to competitive technologies or for recombination of organizational resources. In contrast, firms
operating in a low-tech industry are in a relatively more stable environment and tend to use acquisitions to refine their current technologies (Porrini, 2004).
Due to the digital revolution happening in the last years not only firms operating in high-tech industries are looking for different ways to innovate; firms in the low-tech industries are also forced to become more innovative in order to not face decline or extinction. In the period of 2000 till 2006 more than 55% of the acquisitions made by acquirers operating in a low-tech industry were technological acquisitions (technological acquisition: target firm had patent activity in the five years before the acquisition was made) and 90% of the acquirer firms had patent activity in the three years prior to the M&A. Thus, the question of how a technological acquisition and technological relatedness influences the post M&A innovation output of an acquirer firm is not only relevant for high tech acquirers but also for low tech acquirers.
This paper examines M&A deals of high-tech and low-tech acquirers that were effected in the period of 2000 till 2006 and tests the effect of technological acquisitions on the post M&A innovative performance of the acquirer firm. In addition, this paper tests whether technological relatedness positively affects the post exploitative innovation performance of a firm and if it negatively or non-significantly affects the post exploratory innovative output of the acquirer firm. Finally, the paper tries to identify whether the R&D intensity of the acquirer firm moderates the relationship between technological acquisitions and the exploratory innovation performance of the firm. By creating two subsets in the dataset, one for high tech acquirers and one for low tech acquirers this thesis will contribute to shed light on the unexplored relationship between technological acquisitions and the post M&A innovative performance in the low-tech industries. Due to the different types of measurement that have been used in previous studies and the mixed empirical results on this topic, the two subsets will also make it possible to compare and contrast the results for the research questions mentioned above for the high-tech
acquirers and those for the low-tech acquirers. This will contribute to complement our current understanding of the role of M&A’s on the innovative performance of a firm.
The paper is organized as follows. Section 2 reviews the relevant literature on the topic of M&A and Innovation. Section 3 provides the theoretical background and states the initial hypotheses concerning the relationship between acquisitions and Innovative performance. Section 4 discusses the data and methods that are used in the study. In section 5 the results and conclusions are presented and in section 6 the discussion is provided.
2. Literature Review
The majority of studies focusing on the impact of M&A’s on the innovative performance of the firm have focused on researching whether engaging in an acquisition may lead to higher innovation. Some of these studies have reported a positive impact of engaging in acquisitions on the innovative performance of firms (Desyllas and Hughes, 2010; Capron and Mitchell, 1998, Karim and Mitchell, 2000).
However, other studies have shown the opposite effect. Negative effects of acquisitions on the innovation output of a firm have been attributed to several reasons. The diversion of managerial commitment from the innovation output, due to the time and energy spent on integration activities (Hitt et al., 1990; Hitt et al., 1996); disruptions to the organizational routines and R&D processes(Ranft and Lord, 2002; Puranam et al.,2006); other authors have attributed these negative effects to technological, organizational and market dissimilarities between the acquirer and target firm (Ahuja and Katila, 2001;Hagedoorn and Duysters, 2002; Cloodt et al., 2006; Chakrabarti et al. 1994).
It is important to note that research in this field has been very fragmented qua measurements used and results. First of all, when looking at the impact of M&A’s on the innovative
performance of a company many scholars have operationalized post M&A innovative performance in different ways, such as the creation of new technologies, products, or processes. For example, in their study Karim and Mitchell, 2000 look at how firms use acquisitions to reconfigure their business resources, by looking at the acquirers and targets product lines post acquisition. Their results support the argument that acquisition activity is a key mechanism by which firms change their mix of business resources.
On the other hand, other studies like Ganzaroli et al, 2016; Hausman et al., 1984; Ahuja and Katila, 2001 have used patent counts to measure the innovative performance output of firms post M&A. In their study Ahuja and Katila (2001) look at the effect of tech and non-tech related acquisitions on the innovative output (patents) of firms in the Chemical industry. This study was the first one to introduce an important distinction between acquisitions in which technology is a component and those where it is not. Based on citations their study is one of the most important studies in this field and their results show that the absolute size of the target knowledge base enhances the subsequent innovation performance of the firm. Additionally, their paper demonstrates that the relatedness of the target and acquirer knowledge bases has a nonlinear impact on the innovation performance of the firm. Finally, they show that non-tech acquisitions do not have a significant effect on subsequent innovation performance.
A more recent paper by Cloodt et al. (2006) extends this research and examines the post-M&A innovative performance of acquiring firms in four major high-tech sectors. In contrast to Ahuja and Katila (2001) their findings conclude that “in a high-tech setting the acquisition of a large absolute knowledge base only contributes to improved innovative performance during the first couple of post-M&A years”. Their findings also show a negative relationship between non-technological M&As and the acquirer post innovation performance. They identify a curvilinear impact on the acquiring firm’s innovative performance in regard to the relatedness between the
acquired and acquiring firms’ knowledge bases.
Both of the studies mentioned above only focus on the total innovation performance of a firm. In his paper Wagner (2011) is one of the first ones to investigate the differential effects of tech-related vs non-tech tech-related M&As on the post M&A exploratory and exploitative innovation output of the acquirer firm. His study confirms the important role that technology related acquisitions play on enhancing the exploratory innovation of the firm. This study finds a non-significantly relationship between technology-related acquisitions in the previous three years before the M&A and the total R&D output. In contrast to other studies, this study shows a positive significant relationship between the accumulated non-technological acquisitions and the total R&D output. However, Wagner’s study only focuses on a major high technology sector: the semiconductor industry.
On another note, Porrini 2004, investigates whether acquirers’ and targets’ alliance experience is beneficial to value creation in high-tech and low-tech acquisitions. High-tech acquisitions are acquisitions made by high-tech firms and low-tech acquisitions are acquisitions made by low-tech firms, regardless of whether their targets are high-tech or low-tech (Porrini, 2004). This study shows different results regarding the above-mentioned relationship for high tech and low-tech acquirers. The author appoints that “some researchers that have combined acquisitions by high-tech and low-tech acquirers may have obtained results that have masked effects unique to high-tech acquirers and low-tech acquirers”.
Previous literature researching technological M&A’s and the innovation relationship have mostly focused on high-tech acquisitions. So far, the relationship between technological acquisitions and the post innovative performance of the firm in the low-tech industries has remained unexplored. Specially the differential effects between high tech acquisitions and low-tech acquisitions.
Additionally, the studies that have made the distinction between technological and non-technological M&A’s have mostly only focused on single industries such as pharmaceutical, bio-medical industry and semiconductor industry (e.g. Ganzaroli et al. 2016; Puranam and Srikanth, 2007; Wagner, 2011). Additionally, most of the studies that focus on more than one industry have not made a distinction in the innovative output post M&A, such as exploratory and exploitative performance output with exception of Wagner (2011) and Ganzaroli et al, (2006).
The studies mentioned above have enhanced our understanding of the relationship between
technological M&A’s and innovation performance. The aim of this paper is to build on the
papers mentioned above, by incorporating previous measurements and distinctions in one study to give a clearer perspective on the impact of M&A’s on the innovative performance of a firm. This will be the first study to research the differential effects of high-tech and low-tech
acquisitions, making the distinction between technological and non-technological M&A’s (if
the target involved in the transaction had patent activity in the 5 years prior to the M&A, the
M&A is considered a technological M&A) and also making a distinction between the
explorative and exploitative output. Secondly, in this study a new created measurement of
technological relatedness will be used to identify to what extent the similarity and complementarity between the target and acquirer knowledge bases have on the exploratory and exploitative output of the firm.
This study will also shed light on to the moderating effect of acquirers R&D intensity on tech acquisitions and the exploratory and exploitative innovation output of the firm. Since R&D investments by acquirers could lead to superior innovation outcomes on their own, and could also build absorptive capacity, enabling successful utilization of external sources of knowledge (Ahuja and Katila, 2001; Cohen and Levinthal, 1990).
Size proxies (Mowery et al., 1996) have previously been used to argue that older/ larger firms have high absorptive capacity because they are likely to have accumulated knowledge and developed routines and processes that facilitate assimilation and innovation (Ahuja and Novelli, 2014). In this study size quartiles are created and will be used as control variables in this study.
In conclusion, this study will be able to confirm or reject some findings of previous studies, but it will also generate important new insights related to the differences in innovative performance in high tech and low-tech acquisitions on non-technological and technological related M&A’s taking in consideration technological relatedness and the moderating role of the acquirer firm R&D intensity.
3. Research Question and Hypotheses
This paper focuses on the differential effects of acquisitions in the low-tech and high-tech industries and the relationship between the number of tech acquisitions and the post innovative output of the firm after an M&A. There are two main concepts that need to be highlighted in order to avoid confusion. These concepts are essential to understand the relationships that will be explained in this section. The remaining concepts are explained in section 4.
High-tech acquisitions are acquisitions made by high-tech firms and low-tech acquisitions are acquisitions made by low-tech firms, regardless of whether their targets are high-tech or low-tech (Porrini, 2004). So, when referring to a high-tech acquisition, this study will refer to the industry in which the acquirer firm is active, without taking in consideration the industry of the target firm. In this study two subsets are created when analyzing the data, one for the low-tech acquirers and one for the high-tech acquirers. The reason for taking an emphasis on the acquirer’s characteristics is in line with previous research that argues that post acquisition decisions, tend to be taken by the acquirer’s departments, including the decisions regarding
innovations (Zollo and Singh, 2004). In their study Hambrick and Cannella (1993) indicate that by providing the target firm with their management tools, the acquirer firm tends to colonize the target firm. In this study, we take the “acquirer’s side” only in regard to the industry where the acquirers are active in and when creating the size quartiles.
The second essential concept that will be used in this section are technological and non-technological acquisitions. An M&A is considered to be a technological acquisition if the target firm had any patenting activity during the five years preceding the acquisition. M&As that did not have any patenting activity will be considered as non-technological M&As. So, when referring to a technological M&A we refer to the “technological” characteristics of the target firm acquired.
There are in total four hypotheses that are being tested in this paper. First, I test whether the number of tech acquisitions is positively associated with the total innovative performance of a firm. Second, I test what the effect is of technological relatedness on the post exploratory and exploitative innovative output of a firm. Finally, I expect that the R&D intensity of the acquirer has a moderating effect on the relationship between the number of tech acquisitions and the total innovative output of the firm. The conceptual model is presented below. Each hypothesis and its theoretical background is individually discussed in section 3.2.
Figure 1: Conceptual Models
3.2 Theoretical background and hypotheses
Firms undertake M&A’s for a variety of reasons, sometimes managers use acquisitions as a substitute for innovation because of the risk in pursuing innovation and trade-offs in resource allocations (Hitt et al, 1990).
Based on the knowledge based view and the resource-based theory of the firm, dissimilar knowledge sources are the reason for differences in innovative performance between firms (Bierly and Chakrabarti,1996). So, an M&A has the potential to expand the innovation output of the acquiring firm, by unifying the knowledge bases of both the acquirer and the acquired firm. This unification may provide economies of scale and scope in research and enhance the acquirer’s potential for inventive recombination (Henderson and Cockbur, 1996; Flemming,
H1 Number of Tech Acquisitions Post Innovative Output Technological Relatedness Post Exploitative Innovative Output Post Exploratory Innovative Output H2 H3 R&D Intensity H2 Technological Relatedness
1999; Ahuja and Katila, 2001).
Previous literature, has looked at the impact of technological and/or non-technological acquisitions on the post M&A performance of the acquiring firm in the high-tech industries. Technological M&As enable acquirers to gain access to external sources of innovation (Arora and Gambardella, 1990; Graebner, 2004; Hitt et al., 1996; Ganzaroli et al.,2014), develop and extend their resources and capabilities (Uhlenbruck et al., 2006; Vermeulen and Barkema, 2001; Ganzaroli et al.,2014). Many argue that non-technological acquisitions may be motivated by short term profit growth, market-entry and market-structure related considerations, or by the desire to expand the firm’s product range internationally (Berkovitch and Narayanan, 1993; Chakrabarti et al., 1994; Hagedoorn and Sadowski, 1999; Trautwein, 1990; Cloodt et al., 2001). Since this type of acquisitions can require extensive managerial attention they may lead to a lower managerial commitment to long-term investments in innovation, which leads to a lower innovative output post M&A (Hitt et al., 1996). Results of prior research studying this relationship have been really segmented. Studies such as the one of Cloodt et al 2006 and Wagner 2011 have found evidence that non-technological M&As have a negative impact on the post-M&A innovative performance. In contrast Ahuja and Katila 2001 have found that non-technological M&A’s are non-significant related to the post M&A innovative output. In his study Wagner looked specifically at the effect of tech-acquisitions on the post M&A innovative performance, where he found a non-significant relationship. Since prior studies have only focused on specific high-tech industries, the differential effects between high-tech industries and low-tech industries are interesting to look at. A low-tech firm is expected to have on average less patents than a high-tech firm. This is due to the more stable environment where they operate compared to a high-tech firm. If the low-tech firm makes a tech acquisition it is expected to relatively have a stronger impact on its innovative output post-M&A compared to a high-tech firm that on average has a much higher innovative output. In summary, based on
the empirical results and theoretical arguments discussed above the following hypothesis is proposed:
Hypothesis 1: The number of tech acquisitions is positively associated with the total innovative performance of a firm; this effect is expected to be higher for acquirers in the Low-Tech Industries
Many authors have also looked at the impact of absorptive capacity on the total innovative output of the firm (Ganzaroli et al., 2014; Ahuja and Katila; 2001). The term absorptive capacity was first introduced by Levinthal and March in their 1989 paper and it refers to “one of a firm's fundamental learning processes: its ability to identify, assimilate, and exploit knowledge from the environment”. In previous studies, it is argued that higher R&D investments by acquirer firms could lead to superior innovation outcomes on their own, and could also build absorptive capacity, enabling the successful utilization of external sources of knowledge (Ahuja and Katila, 2001; Cohen and Levinthal, 1990; Srikanth and Puranam; 2007). In their study Ahuja and Katila (2001) argue that a tech acquisition increases the firm knowledge base and therefore has a positive impact on the absorptive capacity of the acquirer’s firm. The majority of the empirical studies on the topic of absorptive capacity, have operationalized absorptive capacity as the R&D intensity of a firm (Meeus et al, 2001; Mowery et al., 1996; Lane et al., 2006). However, most of the studies researching the impact of a M&A on the post innovative output have used R&D as a control variable and the potential moderating effects have not been studied so far. Therefore, hypothesis 2 can be stated as follow:
Hypothesis 2: The R&D intensity of the acquirer moderates the relationship between the number of tech acquisitions and the total innovative output of the firm
knowledge bases is critical for the success of an M&A, if not enough attention is given to this it could be the main reason for failure (Deutsch and West, 2010). Technological overlap between two firms has also been proven to have a positive effect on transaction incidence (Bena and Kai Li, 2014). Many authors have argued that the acquisition of related businesses makes it easier to absorb and integrate the new knowledge into the knowledge base of the acquirer firm. This gives the acquirer firm more slack to invest its resources into innovation activities, which results in higher post M&A innovative output. Authors have made the distinction between complementarity and similarity in the knowledge bases between the acquirer and the target firm. Larsson and Finkelstein (1999) identify two firms to have complementary knowledge bases if they focus on “different narrowly defined areas of knowledge within a broadly define area of shared knowledge”. This implies that the acquirer firm and the target firm share basic understanding of the other firm knowledge base, leading to a higher absorptive capacity of the acquiring firm than in the case of completely unrelated knowledge bases (Ganzaroli et al, 2014). Makri et al (2010) suggests that technology similarity supports exploitation and technological complementarity strengthens exploration. In the case of a related M&A, where the target and the acquirer firm are assigned the same technological codes, it is expected that the acquirer will pursue more related exploitative innovation post M&A. However, in the case of an unrelated M&A, the acquirer gets access to a knowledge base that was previously inaccessible for the firm, and this may enhance explorative innovation (Gupta et al.,2006). Based on the previous literature there is no base to assume that these effects will be different for acquirers in Low-tech industries or High-tech industries. This leads to the following hypotheses:
Hypothesis 3: Technological Relatedness positively affects the post exploitative innovative performance of a firm; this effect is expected to be more or less equal for both Low-Tech and High-Tech acquirers
Hypothesis 4: Technological Relatedness is non-significantly or negatively associated with the post exploratory innovative performance; this effect is expected to be more or less equal for both Low-Tech and High-Tech acquirers
4. Data and Methods
4.1. Model
To analyze the data and test the hypotheses this study uses a panel dataset model that combines time series and cross-sections. In order to analyze the data a negative binomial regression model is used. There are several reasons why this model is used. First, the dependent variables are count variables that take on discrete non-negative integer values, including zero, so a Poisson or a negative binomial specification is recommended (Ahuja and Katila., 2001; Ganzaroli et al., 2015; Hausman et al., 1984; Henderson and Cockburn, 1996). Second, a negative binomial model is chosen over a Poisson regression due to the fact that this model relaxes the restrictive assumption of the mean and variance being equal to each other and it accounts for omitted variable bias while estimating for heterogeneity (Rothaermel and Boeker, 2008). The Hausman’s specification test was used to determine whether to employ fixed effects or random effects, the results showed that random effects would not adequately account for the firm effects and time effects in the data. Therefore, all of the analyses were performed in stata using fixed effects models.
In this study, there are three dependent variables: total innovation output (hypothesis 1 and 4), exploratory innovation output (hypothesis 3) and exploitative innovation output (hypothesis 2). These three dependent variables are non-negative integer–valued count variables that are used as a measure of post-M&A innovative performance. They are measured by the number of patents achieved by the firm in the three years before the M&A (year 0) and the three years after the M&A was effective (year 1, 2 and 3). Lagged effects are included to test the effect of
the acquisitions for up to three years after the year the M&A was originally made. In contrast to previous studies that have taken a panel dataset approach to measure the impact of an M&A on the post innovative output of the firm and that take the number of patents filed in the year when the M&A is effective as year 0; this study takes the average of the number of patents achieved by the firm in the three years before the M&A as the innovative output for year 0. The reason for doing this is to get a more accurate view of the effect of an M&A on the innovative output of the firm post M&A in comparison to the prior years before the M&A. By taking the average of the three years before the M&A for year 0 of the analysis, this study reaches a more accurate view of the effect of an M&A on the post innovative output of the firm. For year 1, 2 and 3 the actual number of patents achieved by the firm for the respective years post M&A are taken. The following model was applied:
In the negative binomial model above exp (εi) Γ[1, α] is assumed to γ distribution. In this formulation of the negative binomial model, the parameter α is estimated directly from the data and captures over dispersion.
4.2 Data and Samples
The hypotheses are tested on an international sample of public firms in the high-tech and low- tech industries, that underwent an M&A in the period between 2000 and 2006. The final sample consists of 299 companies from USA, Europe and Asia. A filter for M&A’s with a completed status was used in order to control for failed M&A’s in the sample. Additionally, minority stake purchases were excluded from the sample. The sample was taken from the Thomson one database, which contains information on the year an M&A was established, the status of the deal (completed or not), acquirer, the target, the parent acquirer and the parent target firm. This database also provides industry information in SIC codes of the target and acquirer. Acquiring
firms were selected based on the industry information provided in SIC-codes, which included industries from the high tech and low-tech industries. The high-tech industries, classified as such by Thomson One, include: communications, computer and office equipment, drugs, electronics and equipment, medical equipment, and computer software. The low-tech industries include: food, tobacco, textile and apparel, wood, paper, chemicals, soaps, rubber, leather, and metals. Companies for which the industry was not mentioned had to be excluded from the initial sample. Additionally, companies that were acquirers and were acquired during the period of 2000 and 2006, were also excluded from the sample. The final sample consist of the following high-tech industries: Bio Technology, Computer equipment, Electronics, Communications and General Tech. The Low-tech sample consist of all of the above mentioned low tech industries. The sample also shows a great deal of variety in terms of the distribution of the size of companies. About 25% of the companies in the sample are relatively small with less than 1500 employees and more than half of the sample (55%) can be found in large size-classes with more than 50,000 employees.
The firms included in the sample do not have the same starting point since it is based on the M&A occurrences in the period of 2000 to 2006. Some firms had more than one acquisition in a year or/and during the period from 2000 to 2006. Based on the effective M&A date of each acquisition, the three years post M&A innovative performance output and other financial and non-financial metrics of each of the acquirers is explored for year 0, 1, 2 and 3. That’s why this sample can still be classified as a balanced panel dataset. The dataset of this study consists of in total 1196 observations (299x4).
In total three different databases were merged to create the final sample of the M&A deals with all of the relevant information that was required. Since the Thomson One database, the CRSP database and the US Patents office database have different identifiers for the firms, a database
that was already used for the paper of technological innovation, resource allocation and growth by Amitt Seru et al. 2017 was used to enable the merging of the three databases. This database made it possible to merge the three databases since it contained information on successful filed patents from 1929 till 2010 extracted from the US Patent Office and Trademark Office database (US Department of Commerce). This database contained more than one million patent applications and each patent was already assigned to their respective company via the permno (company identifier). This means that U.S. patent data is used for all the firms in the sample, including foreign firms. As appointed in previous studies this is necessary to maintain consistency, reliability, and comparability, as patenting systems across nations differ in their application of standards, system of granting patents, and value of protection granted (Ahuja and Katila, 2001). Since the Thomsom one database uses Cusips as the companies’ identifiers and not the permno, in order to find the permno of the respective companies this database needed to be merged with the CRSP/Compustat database first. The CRSP/Compustat database uses the NCUSIP as the companies’ identifiers. After having converted the CUSIPS into the NCUSIPS, it was then possible to find the respective PERMNO of each of the companies. Companies for which a match for the NCUSIP, CUSIP and PERMNO was not found were excluded from the sample. Other financial information such as R&D expenses and sales of the acquirer, and size of the company was extracted from Compustat and CRSP databases.
4.3 Variables
4.3.1 Independent Variables Number of tech acquisitions
In keeping up with literature M&As are reported as technological acquisitions if the target firm had any patenting activity during the five years preceding the acquisition. M&As that did not meet the above-mentioned criterion are considered as non-technological M&As. To distinguish non-technological acquisitions from technological acquisitions, the patenting activity of the
acquired firm in the five years preceding the M&A event will be analyzed. The five-year timeframe to measure the level of technological knowledge has been used frequently before (e.g. Hoetker, 2005; Cloodt et al., 2006; Wagner, 2011); it is considered to provide a suitable balance with regard to addressing the depreciation of the value of technological knowledge of a firm. All of the acquisitions were first tagged with a dummy variable as a technological or non-technological acquisition based on the patents acquired. After this, the number of technological acquisitions in the period between 2000 and 2006 for every specific acquirer were counted. The number of technological acquisitions was taken in consideration for all of the time lags for each specific acquirer during the whole period of the study.
Technological Relatedness
In previous literature, the technological relatedness/distance between firms has been operationalized in different ways. Some authors have measured it based on the overlap of technology codes between acquirers and acquired firms (Seth,1990; Singh and Montgomery,1987; Puranam and Srikanth,2007), while others have looked at the extent to which the acquirer and acquired firm patented in the same three-digit subclasses (Ganzaroli et al.,2014; Diestre and Rajagopalan, 2012; Makri et al., 2010; Rosenkopf and Almeida, 2003; Ahuja and Katila, 2001). This study introduces a measure for technological relatedness between the acquirer and the target firm that is more in line with the operationalization of technological relatedness used in the study of Puranam and Srikanth., 2007. In their study, they measure technological relatedness through the extent of overlap between the technology codes assigned to acquired firms and acquirers by SDC Platinum. They then operationalize the extent of overlap as the number of codes common to the acquirer and acquired firm divided by the total number of technology codes of the acquired firm. The variable to measure the technological relatedness in this study is: Relatedness1. The technological relatedness was
assessed through the extent of overlap between the technology (sic) codes assigned to acquired firms and acquirers by Thomson One. In this study, we diverge somewhat from Puranam and Srikanth in the way that the extent of overlap is measured. By introducing an ordinal scale to this variable, this study tries to capture differences in similarity and complementarity between the target. The scale is as follow:
- Non-related M&A (0): none of the target technology sic codes match with the acquirer sic codes
- Not directly related but fall under the same technology sic code category (1): the technology sic codes of the target do not match specifically with the technological sic codes of the acquirer. For example, if the technological sic codes of one target are 112, 113, 114 and the SIC codes of its Acquirer are 129, 119, 122 even though they do not match exactly they all belong to the category of Biotechnology.
- More or less related (2): some of the target technology sic codes match with the acquirer technology sic codes but the target is also active in other industries in which the acquirer is not active in
- Related M&A (3): all the target technology sic codes match completely with the acquirer technology sic codes
If the acquirer made more than one acquisition in the period of 2000 to 2006, there was more than one target (tech or non-tech) and respectively a different technological (sic) relatedness score for each of the targets. In these type of situations, the average of the technological (sic) relatedness score for the different targets was taken. For example, if an acquirer firm took over two different targets in a year and the relatedness score for target 1 was 2 and the relatedness score of target 2 was 3, the overall relatedness score given to the acquisitions done by that acquirer in year 0 was 2.5. In this way, this study is not only taking into consideration the level of SIC similarity (code category 3 and 2) but also SIC complementarity (code category 1) what
has not yet been done before in prior studies. For a robustness check, a different variable more in line with the previous literature was also used to measure the relatedness, Relatedness2. For this other variable, a categorical variable was used where two of the above-mentioned groups were merged into one. The two categories of this new variable are related or non-related. The two categories of this new variable are as follow: Groups 0 and 1 from the Relatedness1 variable, were identified as a non-related M&A (dummy 0). Group 2 and 3, were identified as a related M&A (dummy 1).
4.3.2 Dependent variables
Patents are directly related to inventiveness (Walker,1995), technological novelty (Griliches, 1990) and have economic significance due to the property rights they confer upon the assignee (Kamien and Schwartz, 1982; Scherer and Ross, 1990; Ahuja and Katila, 2001). Patents have been used to measure the total R&D output/Innovative performance of a firm by several authors e.g. Hausman et al.(1984), Ahuja and Katila, (2001). The specific approach used by Ganzaroli et al. (2016), Wagner (2011) and Cloodt et al. (2006) is used to define the dependent variables of this study. In this study patent counts have not only been used to measure the total innovative output of a firm but they have also been used to build upon and operationalize exploitative and exploratory R&D output.
Patents have been criticized in previous literature arguing the fact that there are some innovations that are not patentable, or are just not patented. In their study Hagedoorn and Cloodt (2003), investigate the innovative performance of a large international sample of nearly 1200 companies in four high-tech industries, using several indicators of innovative performance (ranging from R&D inputs, patent counts and patent citations to new product announcements). Their results suggest that the statistical overlap between these different indicators is so strong that future research might consider just using one of them when
measuring the innovative performance in high tech industries.
Total Innovative output: the number of total patents an acquirer successfully filed in the three-year window post (year 1, year 2 and year 3) and pre-M&A (year 0).
Exploitative innovation output: the number of patents an acquirer successfully filed in the three- year window post-M&A within patent classes in which the firm had been active in the six-year window prior to the M&A.
Explorative innovation performance: the number of patents a firm successfully applied for in the three-year window post-M&A within patent classes in which the firm had not been active in the six-year window prior to the M&A.
4.3.3 Control and moderating variables
Acquirer firm size: Size proxies (Mowery et al., 1996) have previously been used to argue that older/ larger firms have high absorptive capacity because they are likely to have accumulated knowledge and developed routines and processes that facilitate assimilation and innovation (Ahuja and Novelli, 2014). This means that the age and size of the acquirer firms may influence their innovation outcomes (Randt and Lord, 2002). The size of the acquirer firms is determined by the number of employees in the acquired firm and was obtained from Compustat and CRSP. In this study size quartiles are created (refer to table 2) and all of the regressions are run for every size quartile in particular.
R&D intensity: R&D investments by acquirers could lead to superior innovation outcomes on their own, and could also build absorptive capacity, enabling successful utilization of external sources of knowledge (Ahuja and Katila, 2001; Cohen and Levinthal, 1990). The research and development intensity is used as a measure of absorptive capacity and it is calculated by taking the percentage of R&D as of the total sales of the acquirer. This information was obtained from
CRSP and Compustat.
M&A activity level: According to previous literature the more M&A’s a firm undergoes, so the more experience a firm has with doing M&A’s, the easier it will be to successfully combine its current knowledge with the knowledge of the target, which will lead to a higher innovative output. According to Nicholls-Nixon and Woo (2003), the experience of the acquirer in doing M&A’s affects the technical output of the firm, since more experienced firms have the ability to select appropriate partners and learn from them. In this study, we measure the M&A activity level as the number of M&A’s an acquirer did during the period of 2000 and 2006.
5. Results
This section provides the overview of the sample and the main descriptive statistics. Additionally, the regression results and conclusions are also presented in this section.
The descriptive statistics for the data of this research are displayed in Table 1. This table shows the main characteristics of the sample used for this study. As already mentioned in a previous chapter the unit of analysis are the acquisitions made by acquirers in the years of 2000 till 2006. In the sample, there were some firms that completed more than one tech acquisition during the period of the study (2000 -2006). In the three-year lags post acquisition, we accounted for this by taking in consideration how many tech acquisitions a specific firm did in a respective year (cumulative). Table 1 below shows that in total an acquirer made at least one acquisition a year (tech or non-tech) and maximum eight acquisitions in a year (M&A activity level), for which for some acquirers all of the acquisitions made were tech acquisitions. Some of the acquirers in the sample did not make any tech acquisition in the period of the study. In the sample, there is considerable variance in the size of the acquirer firms. The smallest firm in the sample had 570 employees and the largest firm had 3,5 million employees. In order to control
for the size of the acquirers, size quartiles were created and when running the regressions these size quartiles were used as subsets (section 5.2). Refer to table 2 for the description of the size quartiles. Regarding the innovation output of the acquirer firm post M&A, the maximum number of patents successfully filed by an acquirer in the period of the study was 4211. It is also important to note that some of the firms in the sample did not have any exploratory patents noted in the period of the study. In table 3 below the correlation matrix for all of the dependent and independent variables in the sample are previewed; in this section, only the correlations between the independent variables are discussed in this section. The specific regressions for the respective hypotheses where the specific size quartiles were used as subsets are discussed in section 5.2.
Examining the correlations displayed in Table 3, we note that there is only a significant correlation between the Size of the acquirer and the M&A activity level of the acquirer (0,4074*). This suggests that the larger the acquirer, the more firms it acquired in the period of the study. It can also be noted that the rest of the correlations between the different independent variables do not suggest any concerns about collinearity.
Table 1: Sample descriptive statistics
5.2 Regression Results and Conclusions
In table 4 below the regression results of the negative binomial model with distributed lag analysis for the post- M&A total innovative performance of the sample firms are displayed. In this section, the regressions where the different size quartiles were used as subsets will be discussed. The full model is used to discuss the results.
Hypothesis 1 argues that technological M&As are positively associated with the total innovative output of the acquiring firm and that this effect is expected to be higher for acquirers in the low-tech Industries. For the acquirers in the high-tech industries, the individual coefficients of the variables for the number of technological acquisitions are significant for the three years post M&A. and for all of the three years they show a negative sign. The summed coefficient (-0,5327; p: 0,000***), reflecting the total impact, is negative and significant. It appears that the number of tech acquisitions for the acquirers in the high-tech sector is negatively related with the total innovative output and that this effect gets stronger for every year after the acquisition (Year1: -0,21, Year2: -0,29 Year3: -0,44). In order to control for the acquirer size, size quartiles were created and used as subsets (refer to table 2). For every subset, a regression was run and the results can also be seen in table 1. The regressions show that the negative effect is stronger for acquirers that fell in the lowest size quartiles compared to acquirers in the largest size quartiles. For the first year after the acquisition the negative effect was only significant for acquirers in size quartiles 1 and 2.
In contrast for acquirers in the low-tech industries, the summed coefficient (0,46; p: 0,293), reflecting the total impact, is positive but not significant. Additionally, the individual coefficients of the variables for the number of technological acquisitions are only significant for the second and three years post M&A and they show a negative sign. Since, in our sample there were no low-tech acquirers that fell under the size quartiles one and two there are only
results for size quartiles three and four. When running the regressions and taking the size quartiles of the acquirers in consideration it appears that the number of tech acquisitions for the acquirers in the low-tech sector is negatively related with the total innovative output in the second year and third year of the M&A. However, for the second year there are no significant effects for each specific size quartile. In the third-year post acquisition, the number of tech acquisitions has a negative effect for the acquirers that fall in the fourth size quartile (-0,65; p: 0,000**). When comparing the results for the low-tech and high-tech industries based on the size quartiles of the acquirers, specifically for size quartile four, it can be noted that the negative effect of an M&A on the post innovative output of the firm is stronger for the acquirers in the low-tech industries.
All of the above together does not provide support for hypothesis 1. Firstly, a positive effect was expected between the number of tech acquisitions and the innovative output of the acquirer firm post M&A and the results showed a negative effect instead. In regard to the differences between acquirers in the low-tech and the high-tech industries, there was only one significant difference. This difference implies that the negative effect showed to be stronger for acquirers in the low-tech industry in size quartile four (-0,65; p: 0,037**) compared to the high-tech acquirers that fell under the same size quartile (-0,29; p: 0,000***). However, the latter had a higher significance level 0,01 in comparison with 0,05.
The regression results for hypothesis 2 are presented in table 5. Table 5, shows the regression results for the effect of R&D intensity and the interaction effect of the R&D and the number of tech acquisitions, per size quartile and for the three lagged years post acquisition. Hypothesis 2 argues that the R&D intensity moderates the relationship between the number of tech acquisitions and the total innovative output of the firm. The regression results for this interaction effect are not statistically significant, and therefore do not support hypothesis 2.
Table 4: Regression results number of tech acquisitions and total innovation output in the lagged three years post M&A per size quartile
Table 5: Regression results number of tech acquisitions, R&D intensity and the interaction effect of R&D intensity
The regression results for hypothesis 3 are presented in table 6. Hypothesis 3 argues that technological relatedness positively affects the post exploitative innovative performance of a firm and that this effect is expected to be more or less equal for both low-tech and high-tech acquirers. For the acquirers in the high-tech industries, the summed overall coefficient (-0,1165; p: 0,077*), reflecting the total impact of relatedness on the post exploitative innovative output of the firm, is negative and significant. It appears that the relatedness in a tech acquisition for the acquirers in the high-tech sector is negatively related with the total post exploitative innovative output. However, when looking at the individual coefficients of the variables for the technological relatedness for each specific year and size quartile the overall coefficient is only significant for the first-year post M&A (-0,2169, p: 0,002***) and only for the acquirers in the fourth size quartile (-0,4309, p: 0,002***). For the second and third year the overall coefficients are also negative, however not statistically significant. For the second and third year post M&A, only the coefficients for relatedness for acquirers in the fourth size quartile are significant (year 2: -0,31 p: 0,061*; year 3: -0,35 p: 0,034**).
In contrast for acquirers in the low-tech industries, the summed coefficient (0,01; p: 0,952), reflecting the total impact, is positive but not significant. The individual coefficients of the variables for technological relatedness are only significant for the second-year post M&A and it shows a positive sign (year 2: 0,4168, p: 0,099*). Since, in our sample there were no low-tech acquirers that fell under the size quartiles one and two there are only results for size quartiles three and four. When running the regressions and taking the size quartiles of the acquirers in consideration it appears that even though the general regressions for year 1 and 3 independently were positive but not significant the regressions for the relatedness for acquirers in the fourth size quartile are positive and significant (year 2: 0,75 p: 0,016**; year 3: 0,69 p:0,03**). The technological relatedness for the acquirers in the low-tech sector is positively related with the exploitative output post M&A for the three years post M&A if the acquirers
fall in the fourth size quartile. When comparing the results for the low-tech and high-tech industries based on the size quartiles of the acquirers, specifically for size quartile four, it can be noted that technological relatedness has a negative impact for acquirers in the high-tech sector, however it has a positive impact for acquirers in the low-tech sector.
All of the above together does not provide support for hypothesis 3. Firstly, a positive effect was expected between the technological relatedness and the exploitative innovative output of the acquirer firm post M&A and the results showed a negative effect for high-tech acquirers (size quartile 4) and a positive effect for low-tech acquirers (size quartile 4).
For both, the low-tech sector and the high-tech sector a robustness check was performed. In this robustness check instead of using an ordinal variable to measure the level of technological relatedness a dummy was used for related (1) and unrelated (0) acquisitions. The results of the robustness check can also be found in table 6. The regressions for the robustness check show a non-significant effect for relatedness on the post exploitative innovative output for acquirers in the high-tech sector, however they show an even stronger positive effect of relatedness (compared to the ordinal scale) on the post exploitative innovative output of the firm for acquirers in the low-tech sector that fall in the fourth size quartile.
The regression results for hypothesis 4 are presented in table 7. Hypothesis 4 argues that technological relatedness negatively or non-significantly affects the post explorative innovative performance of a firm and that this effect is expected to be more or less equal for both low-tech and high-tech acquirers. No significant results were found for acquirers in the low-tech industry nor for acquirers in the high-tech industry. The results of the robustness check can also be found in table 7. The regressions for the robustness check also show a non-significant effect for relatedness on the post explorative innovative output for acquirers in the high-tech sector and low-tech sector. This means that hypothesis 4 is supported.
Table 6: Regression results number of tech acquisitions and total exploitative innovation output in the lagged three years post M&A per size quartile of the acquirer
Table 7: Regression results number of tech acquisitions and total explorative innovation output in the lagged three years post M&A per size quartile of the acquirer
6. Discussion
This paper explored the differential effects between high-tech and low-tech M&A’s and the impact of technology related M&A’s, technological relatedness and research and development intensity on the total post innovative output, exploitative output and explorative output of the acquiring firm. This paper questions some of the previous findings from single sector studies such as the study of Cloodt et al. (2006), Ahuja and Katila (2001) and also the multi-sectoral study of Wagner (2011), which only focused on high-tech industries.
There are several important insights that the results provide. In contrast to previous studies, the results of this study show that technological M&A’s have a negative impact on the total post innovative performance of a firm (three years post M&A). Additionally, that this negative effect is stronger for smaller acquirers and intensifies for each of the three years post M&A.
As mentioned in the previous chapters, most of previous studies researching the relationship between tech acquisitions and the impact on the innovative output of the firm have implied a positive impact of tech acquisitions on the post innovative output of the firm (Ahuja and Katila, 2001; Cloodt et al., 2006; Wagner, 2011). However, these studies only look specifically at the effect of non-technological acquisitions on the innovative output of the firm where they find a negative effect, except from Wagner (2011). Wagner (2011), looks specifically at the effect of tech acquisitions on the total post innovative output of the firm, specifically in the semiconductor industry. Here, he finds that tech acquisitions non-significantly affect the post innovative output of a firm.
Previous literature on the topic of M&A and innovation, has also shown negative effects of an M&A on the innovative output of a firm (Hitt et al., 1990; Hitt et al., 1996, Desyllas and Hughes, 2010). In this case, the theory of Hitt et al., 1990 may be one potential explanation of the negative relationship, their paper suggests that there is a tradeoff between growth through
acquisitions and managerial commitment to innovation in the acquiring firm. In this study, commitment to innovation is defined as the managerial willingness to allocate resources and champion activities that lead to the development of new products, technologies, and processes consistent with marketplace opportunities. Based on this theory a potential reason for negative effects of an M&A on the post innovative output of a firm is the amount of managerial energy absorbed by the acquisition process. According to this theory, “the absorption of a new firm results in a lower propensity for other managers within the firm to pursue risky projects that require the support of top-level managers whose energies are directed primarily toward the acquisition process”. Other potential reason may be that due to increased debt levels that are normally necessary to fund the acquisitions (Michel and Shaked, 1985), debt holders gain power relative to the other stakeholders; debtholders are usually more risk averse compared to equity holders (Smith and Warner, 1979; Williamson, 1988). Therefore, the acceptability to pursue riskier projects may decrease. Other authors also suggest a negative effect of acquisitions on the innovative performance of the acquirer firm due to the influence of temporary restructuring costs and the disruption of established organizational and R&D routines (Haspeslagh and Jemison, 1991; Ranft and Lord, 2002; Desyllas and Hugues, 2010).
The results of this study also show that technological relatedness has a negative impact on the exploitative innovative output post M&A for acquirers in the high-tech sector, however it has a positive impact for acquirers in the low-tech sector. The contradicting results compared to previous literature on this topic may be due to the broad sample taken compared to the single sector studies and differences in measurements. Secondly, when comparing the results for the low-tech and high-tech industries based on the size quartiles of the acquirers, specifically for size quartile four, the results showed that technological relatedness has a negative impact for acquirers in the high-tech sector, however it has a positive impact for acquirers in the low-tech
sector. In previous studies, the relationship between technological relatedness of the acquirer firm and the target firm and the innovative performance post M&A has been conceptualized in many different ways (Ganzaroli et al. 2016). The most used conceptualization is based on a definition of technological distance as continuous between dissimilarity and similarity (e.g. Nooteboom et al., 2007; Nesta and Saviotti, 2005; Ganzaroli et al. 2016). These studies argue that the relationship between technological “relatedness” and innovation performance is supposed to be shaped as an inverted U. In this case, it means that too much similarity(relatedness) between the target’s and acquirer’s knowledge base bounds the potential for innovative recombination. Other possible explanation for the negative effect of relatedness on the post exploitative innovative output of the firm, may be that when a firm makes a related acquisition, that there may be other reasons for the acquisition that are not tied to the innovative performance of the firm. Examples of other possible reasons for acquisition are entry to another market, economies of scale/scope, product development or process innovation. If after the M&A the acquirer firms put all of their resources to pursue these other goals, this could be a reason for the acquired negative effect on the post exploitative innovative output of the firm. This study also shows a non-significant relationship between technological relatedness and post exploratory innovation output. The findings mentioned above deserve further attention.
The managerial implications of this paper are tied to aspects of the target selection and integration process of the acquisition. This study indicates that the number of technological acquisitions a firm does in a year is associated with negative innovative output in the three years post acquisition, both for acquirers in the high-tech and low-tech industries. Managers of firms whose acquisition goal is to increase their innovation output, have to be aware that this may not happen in the short term. They need to have available resources for the post integration process of the target firm and work towards an efficient and effective integration. Further from
the target selection perspective, technological relatedness is favorable for acquirers in the low-tech industries, for the post exploitative innovation output.
6.1 Limitations and Further Research
There are several limitations that should be taken into account when assessing the main findings of this study. The main limitation of this paper is the relative small sample size and its composition. This paper examines M&A deals of high-tech and low-tech acquirers that were effected in the period of 2000 till 2006, however due to data availability restrictions at the time of matching the three used databases (Thomson One, US patent office database (Amitt Seru database) and CRSP/Compustat) more than 50% of the M&A deals from the original sample needed to be extracted. Future research should try to find other databases that better match with each other to obtain a larger sample.
Future research could also extend this study by studying the post innovative output of the firm in the long term (longer than three years) and considering the size of the target firm (knowledge base). Additionally, instead of only focusing on patents further research could also look at combined measurements of innovative performance such as new products introduced after the M&A by the acquirer firm and process innovations. All of this could give more clarification on the real impact of an M&A on the post innovative performance of a firm.
Reference List
Ahuja, G., and Katila, R. (2001). ‘Technological Acquisitions and the Innovation Performance of Acquiring Firms: A Longitudinal Study’, Strategic Management Journal, 22: 197–220 Ahuja, G., and Novelli, E. (2014). ‘Mergers and Acquisitions and Innovation’, The Oxford Handbook of Innovation Management
Bierly, P., and Chakrabarti, A. (1996). ‘Generic knowledge strategies in the U.S. pharmaceutical industry’, Strategic Management Journal, Vol. 17(Winter Special Issue), 123-135
Ranft, A.L., and Lord, M.D. (2002). ‘Acquiring new technologies and capabilities: a grounded model of acquisition implementation’, Organization Science, 13(4), pp.420-441
Arora, A. and Gambardella, A. (1990), ‘Complementarity and external linkages: the strategies of the large firms in biotechnology’, The Journal of Industrial Economics, Vol. 38 No. 4, pp. 361-379
Seru, A., Kogan, L.,Papanikolaou, D., Stoffman, N. (2016). ‘Technological Innovation,
Resource Allocation, and Growth’. NBER Working Paper No. 17769
Bena, J., and Kai, L. (2014). ‘Corporate Innovations and Mergers and Acquisitions’, The Journal of Finance, 69 (5).
Berkovitch, E., Narayanan, P. (1993). ‘Motives for takeovers: an empirical investigation’. Journal of Financial and Quantitative Analysis 28, 347–362
Capron, L., Dussauge, P., and Mitchell, W. (1998). ‘Resource Redeployment Following Horizontal Acquisitions in Europe and North America, 1988–1992’, Strategic Management Journal, 19(7): 631–61.
Chakrabarti A., Hauschildt J., Suverkrup C. (1994). ‘Does it pay to acquirer technological firms?’, R&D Management 24(1):47-56.
Cloodt, M., Hagedoorn, J., and Van Kranenburg, H. (2006). ‘Mergers and Acquisitions: Their Effect on the Innovative Performance of Companies in High-Tech Industries’, Research Policy, 35: 642–54
Cohen WM., Levinthal D. (1989). ‘Innovation and learning: the two faces of R&D’. Economic Journal 99(397): 569-596
Cohen, W. M., and Levinthal, D. A. (1990). ‘Absorptive Capacity: A New Perspective on Learning and Innovation’, Administrative Sciences Quarterly, 35: 569–96
Franks, J., Harris, R., & Titman, S. (1991). ‘The post-merger share-price performance of acquiring Firms’. Journal of Financial Economics, 29: 81-96.
Desyllas, P., and Hughes, A. (2010). ‘Do High Technology Acquirers Become More Innovative?’, Research Policy, 39: 1105–21
Deutsch, C. and West, A. (2010), ‘Perspectives on Merger Integration’, McKinsey Company, Brussels
Diestre, L. and Rajagopalan, N. (2012), ‘Are all ‘sharks’ dangerous? New biotechnology ventures and partner selection in R&D alliances’, Strategic Management Journal, Vol. 33 No. 10, pp. 1115-1134.
Flemming L. (1999). ‘Recombinant uncertainty in technological search’, Working paper, Harvard Business School.
Ganzaroli, A., De Noni, I., Orsi, L., Bellusi, F. (2016). ‘The combined effect of technological relatedness and knowledge utilization on explorative and exploitative invention performance post-M&A’, European Journal of Innovation Management, 19 (2), 150-156
Graebner, M. (2004). ‘Momentum and serendipity: how acquired leaders create value in the integration of technology firms’, Strategic Management Journal 25, 751–777
Griliches Z. (1990). ‘Patent statistics as economic indicators: a survey’. Journal of Economic Literature 28(4): 1661-170
Gupta, A.K., Smith, K.G., Shalley, C.E. (2006). ‘The interplay between exploration and exploitation’, Academy of Management Journal 49, 693–706.
Hagedoorn, J., Sadowski, B. (1999). ‘The transition from strategic tech- nology alliances to mergers and acquisitions: an exploratory study’, Journal of Management Studies 36, 87–107 Haleblian, J., Devers, C. E., McNamara, G., Carpenter, M. A., & Davison, R. B. (2009). ‘Taking stock of what we know about mergers and acquisitions: A review and research agenda’, Journal of Management, 35(3), 469-502
Hambrick, D., Cannella, A. (1993). ‘Relative standing: a framework for understanding departures of acquired executives’. Academy of Management Journal 36 (4), 733–762
Haspeslagh, P.C., Jemison, D.B. (1991). ‘Managing Acquisitions: Creating Value Through Corporate Renewal’. Free Press, New York.
Hausman J, Hall BH, Griliches Z. (1984). ‘Econometric models for count data with an application to the Patents-R&D relationship’, Econometrica 52(4): 909-938.
